Previsão da Irradiação Solar Fotovoltaica baseada em Deep Learning

  • Felipe de Sena Rodrigues Departamento de Engenharia Mecânica, Universidade Federal da Bahia, Salvador-BA
  • Ângelo Márcio Oliveira Sant’Anna Departamento de Engenharia Mecânica, Universidade Federal da Bahia, Salvador-BA
  • Daniel Barbosa Departamento de Engenharia Elétrica, Universidade Federal da Bahia, Salvador-BA
  • Nyegirton Barreiros dos Santos Costa Instituto Federal do Sertão Pernambucano, PE
Keywords: artificial intelligence, deep learning, forecasting, solar radiation, photovoltaic energy

Abstract

The Northeast area has shown great potential for generating of solar energy in Brazil, for this reason several energy companies have evaluated the economics feasibility this kind of renewable energy. Thus, as it increases the portion of renewables sources in supply of energy is important to estimate your potential with an acceptable accuracy in distincts horizons, like hours, days, weeks, and months. This paper proposed to develop distinct deep learning models for forecasting the global solar irradiation. The measures were collected from a solarimetric station located in the State of Bahia. The models are evaluated through five performance metrics, and the best results were selected for the final models. Finally, Friedman statistical test and Bergmann-Hommel’s post hoc test were applied to evaluate the hypothesis for significance difference between the developed models.
Published
2022-11-30
Section
Articles